Ensuring that software applications presents their users the most recent version of data is not trivial. Self-adjusting computations are a technique for automatically and efficiently recomputing output data whenever some input changes.
This repository contains the code of a framework for coarse-grained self-adjusting computations in Haskell. The framework has been extracted from a commercial software product developed by medilyse GmbH, Freiburg, Germany.
The repository also contains two demo applications using the framework.
There is an accompanying article:
Stefan Wehr. 2023. A Software Architecture Based on Coarse-Grained Self-Adjusting Computations. Proceedings of FUNARCH 2023. Seattle, WA, USA. ACM, 2023. Preprint
Requirements: you need the build tool stack (https://docs.haskellstack.org/en/stable/).
I use version 2.9.3, I guess a slightly older version might work was well.
Build the software by executing
$ stack build
Run the tests by executing
$ stack test
$ stack run -- test
This demo synchronizes the content of one directory to another. The sync works recursively and continuously.
To start the sync from directory A to directory B, execute the following command.
Note that A must exist, B is created automatically.
Attention: everything in B that does not exist in A will be deleted!
$ stack run -- --log-level info sync --source A --target B
This demo shows how a processing pipeline for hospital data might look like. The demo is very basic. It supports admission and discharge of patients, as well textual notes being added to a patient. A simulation creates a stream of events driving the demo.
Start the simulation:
$ stack run -- --log-level info hospital-simulation --root run-data
Start the pipeline:
$ stack run -- --log-level info hospital-pipeline --config hospital-config/ --root run-data
Now output files should appear in run-data/output.
You can also change data by hand. The sqlite database for patients is in
run-data/pats.sqlite, the database for patient notes in
run-data/pat_notes.sqlite.
Start the webserver:
stack run -- --log-level info hospital-server --out run-data/output --web webapp
Then point your browser to http://localhost:8080